Topic: Swedish ASIC miner company kncminer.com - page 1637. (Read 3049501 times)

Using ultrasonic soldering, assembly of flip-chip packages can hit 20,000 per hour.  20K chips hardly warms up the machines - it's a very short order to be running...

So a question to the resident experts: If they are right on schedule and everything goes without a hitch, how long is it going to take them to slice/dice/package 10,000 or 20,000 quad-core processors?

Wasn't intended to be rude, he said it himself that he had never designed an ASIC before.

I think that says a lot about the talent and dedication of the man, actually.

Bitfury 55nm = .9W/GH

How can KnC not be less than that with a process size 50% of Bitfury's chip?  Hopefully the simulations are wrong and KnC will over-deliver.

But I guess this bodes well for Bitfury ver.2.0. 

BFL was a custom ASIC, and all your figures are wrong.

Avalon/AM 0.133GHs/W
BFL  0.157 GHs/Watt
KNC 0.8GHs/Watt
Bitfury 1GHs/Watt

Obviously if you overclock the figures will drop, and KNC say they have understated their performance, so it should be closer to Bitfury performance.

Right, because Bitfury have done nothing?.. except deliver ASICs, on time, that worked as promised and at a reasonable price.

I'm starting to get a Ford/Chevy rivalry vibe coming from the KnC fanclub (of which I'm a proud member).  It's surprising to me how few of the posters in this thread even acknowledge the Bitfury Group.  Is 55nm so beneath everyone?

If nothing else, Bitfury and (hopefully) KnC are head-and-shoulders above what we've been used to dealing with.  And I feel sorry for the late-comers such as Cointerra, HashFast, etc.  They're going to have a very difficult time finding a spot on the mountain next to the new gods of ASIC.

Valid point, bad comparison. Bitfury is playing in it's own league, being full custom ASIC. And we all know BFL didn't have realistic expectations for their chips.

Existing standard cell designs:
110nm = 8W/Gh (AM)
65nm = 4W/Gh (BFL)
28nm = ??

KnC sits on the right spot with their estimate of 1.8W/Gh so if only their chips work they should be good

Calling Bitfury an amateur is pretty rude IMO. Because their 65nm = 0.9W/Gh.

Yeah,it seems it's gonna be an interesting few months and this is the first company to offer some real ASIC competition.

How so?

I'd like it cheaper obviously, but it's actually confirmed working and therefore a better bet to ship on time next month. The cost difference between a bitfury 400Gh rig ordered today and a knc jupiter ordered months ago is negligible, though I realize EU & US prices may differ. 

I'm looking at a possible knc order for November delivery depending on how they handle shipping. If they knock it out the park like bitfury did then it'll be a viable alternative.

But the price makes it prohibitive unfortunately!

KNC mentioned in Life On Bitcoin's latest video.

http://www.youtube.com/v/GMsHpaHP9f0

guess we shall see about that soon enough... 

Amateur...pro who cares. The bitfury chip is better than anything else out there at this time. This fact is not likely to change next month btw.

We of course don't know. However, their recent announcement that the chips do indeed hash faster than advertised would tend to indicate that they had sample chips to work with.

Also, both BFL and Bitfury (regardless of their morals, which seem to be on opposite poles) are essentially amateurs, whereas the KnC/OrSOC team are ASIC engineers. The only thing new to them on this, aside from the venture itself, is the Application that is Specific to the Integrated Circuit. Seasoned professionals generally get better results for the effort involved than amateurs. No offense intended toward Bitfury, as that is one TALENTED amateur.

They will need to have atleast few chips to test the chip itself. They must get a lot data from the actual chip - how can they ship a product to end-user that includes a chip that has only been simulated?
If they don't have chips to test yet, they will be late. Just like Bitfurys chip differed from the simulations kinda lot. Same with BFL. How would it be different with KNC? Really?

you missed the part where bfl had the chips pretty much paid for in preorder money when they started to produce them and where they decided to part with the old chips and try out their "next gen" miner...
That is pretty much the difference, KNC needed the preorder money to fund the masks, BFL had them already. Maybe you will see another payment mode on the gen 2 models...

I apologize profusely to those who have been waiting for delivery from BFL, in some cases for over a year, and I realize this may be adding salt to the wound but in the interest of full disclosure they appear to be sending our bulk chip orders on schedule. I received my chips exactly 100 days after paying my deposit. I bought from KnCminer and Metabank at the same time and have received nothing so far from them. Not sure whether this is because it's much simpler to send just chips (Avalon chip buyers might disagree) or because BFL was motivated to collect the balance due on delivery. However this might be a better pre-order model going forward: Collect half upfront and half on delivery...

There was a question about the feasability of solo mining.
Go here: www.bitcoinx.com/profit/
Punch in your hashrate and expected difficulty and read out your chances.

I completely agree with you that it is better for the miner when it works / the pool is well funded and reputable. The thing to lookout for is when everyone starts piling into the pool and the luck doesn't adjust to the newly recruited miners... with the amount of interest/hashpower that its being brought online, the risk increases.

How sure are you that they are not using PPS? I originally created an account with them until I read some wheres they were a PPS pool then the buck stopped there.

That is a half truth. IF the pool has enough BTC in reserve, it will pay you out regardless of whether the shares you contributed amounted to solving a block or not. now, IF the pool DOES NOT have enough BTC in reserve, and you contribute X amount of shares towards a block which amounts to nothing, then the pool has an obligation (and I use the term obligation loosely) to pay you out for each share even if the pool did not solve a block. Then the owner of the pool has to make a decision. A. Pay out of his own pocket the BTC owed to you, OR B. Bankrupt the Pool and leave you high and dry....It has happened. research Bitclockers.com - the pool tried pay PPS at a less than 3% rake and the pool went bankrupt. alot of people lost alot of coin because of this. I personally lost .29 BTC but it was enough to make me pay attention to how a pool operates much more acutely.

PPLNS is different as it will never pay out more than the pool brings in. Ultimately, its up to you, if you wanna risk it with PPS, and you feel comfortable with it, go ballz deep.. no problem. Just know what your getting into.

On a side note, One lesson learned by a lot of people over at bitclockers is that anyone who is mining large amounts of coin should not leave it inside the pool. Set your auto payout triggers to something low so that you will know IF and/or a pool is having a problem paying its miners.